End of training

README.md

@@ -1,4 +1,6 @@
 ---
+language:
+- en
 license: apache-2.0
 tags:
 - generated_from_trainer
@@ -13,7 +15,7 @@ model-index:
       name: Text Classification
       type: text-classification
     dataset:
-      name: glue
+      name: GLUE MNLI
       type: glue
       config: mnli
       split: validation_matched
@@ -21,7 +23,7 @@ model-index:
     metrics:
     - name: Accuracy
       type: accuracy
-      value: 0.6193581253183902
+      value: 0.6172701383238405
 ---
 
 <!-- This model card has been generated automatically according to the information the Trainer had access to. You
@@ -29,10 +31,10 @@ should probably proofread and complete it, then remove this comment. -->
 
 # mobilebert_sa_GLUE_Experiment_logit_kd_mnli
 
-This model is a fine-tuned version of [google/mobilebert-uncased](https://huggingface.co/google/mobilebert-uncased) on the glue dataset.
+This model is a fine-tuned version of [google/mobilebert-uncased](https://huggingface.co/google/mobilebert-uncased) on the GLUE MNLI dataset.
 It achieves the following results on the evaluation set:
-- Loss: 1.2093
-- Accuracy: 0.6194
+- Loss: 1.1966
+- Accuracy: 0.6173
 
 ## Model description
 

all_results.json

@@ -0,0 +1,21 @@
+{
+    "epoch": 15.0,
+    "epoch_mm": 15.0,
+    "eval_accuracy": 0.6249617931737137,
+    "eval_accuracy_mm": 0.6172701383238405,
+    "eval_loss": 1.1755008697509766,
+    "eval_loss_mm": 1.1966100931167603,
+    "eval_runtime": 23.0085,
+    "eval_runtime_mm": 23.0079,
+    "eval_samples": 9815,
+    "eval_samples_mm": 9832,
+    "eval_samples_per_second": 426.582,
+    "eval_samples_per_second_mm": 427.331,
+    "eval_steps_per_second": 3.347,
+    "eval_steps_per_second_mm": 3.347,
+    "train_loss": 1.177346492735835,
+    "train_runtime": 35549.2544,
+    "train_samples": 392702,
+    "train_samples_per_second": 552.335,
+    "train_steps_per_second": 4.315
+}

eval_results.json

@@ -0,0 +1,16 @@
+{
+    "epoch": 15.0,
+    "epoch_mm": 15.0,
+    "eval_accuracy": 0.6249617931737137,
+    "eval_accuracy_mm": 0.6172701383238405,
+    "eval_loss": 1.1755008697509766,
+    "eval_loss_mm": 1.1966100931167603,
+    "eval_runtime": 23.0085,
+    "eval_runtime_mm": 23.0079,
+    "eval_samples": 9815,
+    "eval_samples_mm": 9832,
+    "eval_samples_per_second": 426.582,
+    "eval_samples_per_second_mm": 427.331,
+    "eval_steps_per_second": 3.347,
+    "eval_steps_per_second_mm": 3.347
+}

logs/events.out.tfevents.1675112283.serv-3330.1319687.26

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:d7a9a810c47caf54aba7d0bce7dca91577e0c7bb9c8a4d84f1999349cde4fd1a
+size 698

train_results.json

@@ -0,0 +1,8 @@
+{
+    "epoch": 15.0,
+    "train_loss": 1.177346492735835,
+    "train_runtime": 35549.2544,
+    "train_samples": 392702,
+    "train_samples_per_second": 552.335,
+    "train_steps_per_second": 4.315
+}

trainer_state.json

@@ -0,0 +1,250 @@
+{
+  "best_metric": 1.1755008697509766,
+  "best_model_checkpoint": "mobilebert_sa_GLUE_Experiment_logit_kd_mnli/checkpoint-30680",
+  "epoch": 15.0,
+  "global_step": 46020,
+  "is_hyper_param_search": false,
+  "is_local_process_zero": true,
+  "is_world_process_zero": true,
+  "log_history": [
+    {
+      "epoch": 1.0,
+      "learning_rate": 4.9e-05,
+      "loss": 1.6232,
+      "step": 3068
+    },
+    {
+      "epoch": 1.0,
+      "eval_accuracy": 0.5504839531329597,
+      "eval_loss": 1.3869810104370117,
+      "eval_runtime": 22.8066,
+      "eval_samples_per_second": 430.358,
+      "eval_steps_per_second": 3.376,
+      "step": 3068
+    },
+    {
+      "epoch": 2.0,
+      "learning_rate": 4.8e-05,
+      "loss": 1.4341,
+      "step": 6136
+    },
+    {
+      "epoch": 2.0,
+      "eval_accuracy": 0.5833927661742231,
+      "eval_loss": 1.3186262845993042,
+      "eval_runtime": 22.7427,
+      "eval_samples_per_second": 431.567,
+      "eval_steps_per_second": 3.386,
+      "step": 6136
+    },
+    {
+      "epoch": 3.0,
+      "learning_rate": 4.7e-05,
+      "loss": 1.3724,
+      "step": 9204
+    },
+    {
+      "epoch": 3.0,
+      "eval_accuracy": 0.5942944472745797,
+      "eval_loss": 1.2818658351898193,
+      "eval_runtime": 22.8582,
+      "eval_samples_per_second": 429.386,
+      "eval_steps_per_second": 3.369,
+      "step": 9204
+    },
+    {
+      "epoch": 4.0,
+      "learning_rate": 4.600000000000001e-05,
+      "loss": 1.3249,
+      "step": 12272
+    },
+    {
+      "epoch": 4.0,
+      "eval_accuracy": 0.5981660723382578,
+      "eval_loss": 1.270175814628601,
+      "eval_runtime": 22.8992,
+      "eval_samples_per_second": 428.618,
+      "eval_steps_per_second": 3.363,
+      "step": 12272
+    },
+    {
+      "epoch": 5.0,
+      "learning_rate": 4.5e-05,
+      "loss": 1.2788,
+      "step": 15340
+    },
+    {
+      "epoch": 5.0,
+      "eval_accuracy": 0.6030565461029037,
+      "eval_loss": 1.235905408859253,
+      "eval_runtime": 22.929,
+      "eval_samples_per_second": 428.06,
+      "eval_steps_per_second": 3.358,
+      "step": 15340
+    },
+    {
+      "epoch": 6.0,
+      "learning_rate": 4.4000000000000006e-05,
+      "loss": 1.2302,
+      "step": 18408
+    },
+    {
+      "epoch": 6.0,
+      "eval_accuracy": 0.6192562404482934,
+      "eval_loss": 1.200829029083252,
+      "eval_runtime": 22.985,
+      "eval_samples_per_second": 427.017,
+      "eval_steps_per_second": 3.35,
+      "step": 18408
+    },
+    {
+      "epoch": 7.0,
+      "learning_rate": 4.3e-05,
+      "loss": 1.1842,
+      "step": 21476
+    },
+    {
+      "epoch": 7.0,
+      "eval_accuracy": 0.6222109016811004,
+      "eval_loss": 1.1990573406219482,
+      "eval_runtime": 22.7719,
+      "eval_samples_per_second": 431.015,
+      "eval_steps_per_second": 3.381,
+      "step": 21476
+    },
+    {
+      "epoch": 8.0,
+      "learning_rate": 4.2e-05,
+      "loss": 1.1441,
+      "step": 24544
+    },
+    {
+      "epoch": 8.0,
+      "eval_accuracy": 0.6201732042791646,
+      "eval_loss": 1.1838924884796143,
+      "eval_runtime": 22.9011,
+      "eval_samples_per_second": 428.582,
+      "eval_steps_per_second": 3.362,
+      "step": 24544
+    },
+    {
+      "epoch": 9.0,
+      "learning_rate": 4.1e-05,
+      "loss": 1.1057,
+      "step": 27612
+    },
+    {
+      "epoch": 9.0,
+      "eval_accuracy": 0.624350483953133,
+      "eval_loss": 1.1861381530761719,
+      "eval_runtime": 22.9564,
+      "eval_samples_per_second": 427.55,
+      "eval_steps_per_second": 3.354,
+      "step": 27612
+    },
+    {
+      "epoch": 10.0,
+      "learning_rate": 4e-05,
+      "loss": 1.0715,
+      "step": 30680
+    },
+    {
+      "epoch": 10.0,
+      "eval_accuracy": 0.6249617931737137,
+      "eval_loss": 1.1755008697509766,
+      "eval_runtime": 22.9046,
+      "eval_samples_per_second": 428.516,
+      "eval_steps_per_second": 3.362,
+      "step": 30680
+    },
+    {
+      "epoch": 11.0,
+      "learning_rate": 3.9000000000000006e-05,
+      "loss": 1.0386,
+      "step": 33748
+    },
+    {
+      "epoch": 11.0,
+      "eval_accuracy": 0.6312786551197147,
+      "eval_loss": 1.1971582174301147,
+      "eval_runtime": 23.1346,
+      "eval_samples_per_second": 424.256,
+      "eval_steps_per_second": 3.328,
+      "step": 33748
+    },
+    {
+      "epoch": 12.0,
+      "learning_rate": 3.8e-05,
+      "loss": 1.0066,
+      "step": 36816
+    },
+    {
+      "epoch": 12.0,
+      "eval_accuracy": 0.6276107997962302,
+      "eval_loss": 1.2148597240447998,
+      "eval_runtime": 22.894,
+      "eval_samples_per_second": 428.714,
+      "eval_steps_per_second": 3.363,
+      "step": 36816
+    },
+    {
+      "epoch": 13.0,
+      "learning_rate": 3.7e-05,
+      "loss": 0.9767,
+      "step": 39884
+    },
+    {
+      "epoch": 13.0,
+      "eval_accuracy": 0.6192562404482934,
+      "eval_loss": 1.2187175750732422,
+      "eval_runtime": 22.8537,
+      "eval_samples_per_second": 429.471,
+      "eval_steps_per_second": 3.369,
+      "step": 39884
+    },
+    {
+      "epoch": 14.0,
+      "learning_rate": 3.6e-05,
+      "loss": 0.9482,
+      "step": 42952
+    },
+    {
+      "epoch": 14.0,
+      "eval_accuracy": 0.6226184411614876,
+      "eval_loss": 1.2003837823867798,
+      "eval_runtime": 23.0327,
+      "eval_samples_per_second": 426.133,
+      "eval_steps_per_second": 3.343,
+      "step": 42952
+    },
+    {
+      "epoch": 15.0,
+      "learning_rate": 3.5e-05,
+      "loss": 0.921,
+      "step": 46020
+    },
+    {
+      "epoch": 15.0,
+      "eval_accuracy": 0.6193581253183902,
+      "eval_loss": 1.2092907428741455,
+      "eval_runtime": 22.9952,
+      "eval_samples_per_second": 426.829,
+      "eval_steps_per_second": 3.349,
+      "step": 46020
+    },
+    {
+      "epoch": 15.0,
+      "step": 46020,
+      "total_flos": 1.8469785301942272e+17,
+      "train_loss": 1.177346492735835,
+      "train_runtime": 35549.2544,
+      "train_samples_per_second": 552.335,
+      "train_steps_per_second": 4.315
+    }
+  ],
+  "max_steps": 153400,
+  "num_train_epochs": 50,
+  "total_flos": 1.8469785301942272e+17,
+  "trial_name": null,
+  "trial_params": null
+}